Tag: howto

I just discovered another interesting use case for current LLM tools. I use Claude.ai at the moment it’s the best in class. I use the free version.

When you design on a document, let’s say a GDD or a one pager for a feature, you should cover the most possible edge cases. In this way, your coders will have everything pretty much clear by reading the document.

If you submit a document to Claude, then you can ask it to simulate the coding in Unity or Unreal Engine for that GDD. It will spit out pseudo code and reasonings. There will probably be lots of hallucinations and things unuseful. Still, there are good chances to discover something you didn’t considered.

When I was 20 in Naples, it happened that I was with friends and we met unknown people. And we discover some party at a student’s house to sneak in. Sometimes it went well, others we would find ourselves with very different people. Among ourselves, sitting in a corner, sipping the same beer we had bought to join the party.

Today I don’t sneak in anywhere anymore. I like parties, but I prefer the comfort of being with people I already know. Sometimes I go to strangers’ houses, but in general I like to know a sufficient number of people.

The same goes for job hunting. I prefer having groups of friends in the industry and moving with them. I have a job right now because during the first months of my daughter’s life I played Fortnite at night and wanted to learn how to make levels. That’s how I met the person who then put me in the loop. This is the healthiest way to learn about realities and find opportunities.

Often it happens that your team is developing what you have defined in your design documentation. It happens that you don’t know specifically what to do next.

I use these moments to play competitor games and engage with their audience. I join Discord servers, lurk subreddits, and so on. I try to keep myself in the shoes of a Player of my references.

The second thing I do is to play the last build or the last code version intensively. As a game designer you need to be aware of what game is coming out.

I also keep ordered single big documents containing everything that has been already defined and implemented in the game. It is very useful to keep track of the vision and its progress.

The biggest Italian books publisher rejected The Lord of The Rings, many years ago. The first reaction I had when I saw this post the other day was from my belly.

But I had the time to reflect: to me, LotR is a masterpiece, but it has also a strong “personality”. It’s not an average story; you may like it or not for specific reasons.

Imagine to be the person in charge of selecting the best books. Something like this arrives, you read it and you like it. But its personality is very strong, and your role requires you to be cold and professional.

Today I can read these words and think that the decision maker made a mistake. But then I put myself in their shoes and it’s hard to make a decision with something like Tolkien’s book. A modern Tolkien can write tales in smaller formats on platforms like Substack. With nobody predicting if the text could be right for a specific geography.

It happens also with video games, I have to say especially to us game designers. Part of our responsibility is to pitch new ideas for features and mechanics. All companies have people in charge that filter our proposals. This process filters also good things out, it’s part of the game.

To design and produce you often need a sort of authorization from people in charge. These people, like you, make mistakes. The lesson you take can be that this is unacceptable, but often it’s a matter of communication. If something so special like The Lord of The Rings arrives on your desk, you can make errors. That’s why people that get a proposal should be prepared first.

I don’t know the full story, maybe the Tolkien’s agent just sent the book with a note. When you present something, the receivers should know all that from before. In that way, they will feel already connected and familiar to the thing.

Like new writers on Substack, leave crumbs and taste the environment!

These days for work reasons I am playing a bunch of simulation games (sims) of different nature. I am making in my spare time a sandbox simulator of a police dog, and then for other reasons I am playing other sims.

They all have a single thing in common, despite belonging to different genres. In fact, immersive sims come from first person shooters, a specific type of action game. Colony sims come from strategy games. Sandbox sims come from adventure games.

They all share a single reason of why Players do love them: they foster the player agency, letting them create their own story.

Everytime I join one of their Discord for research I feel that. Every player has a different story to tell. That is, in my opinion, the single most important motivation for the core audience of sim games. Simulation doesn’t mean simulator, high realism and fidelity (also if some simulation game puts “simulator” in its name).

It means to have many levers to move to create your own story for your friends, Discord, or a livestream.